THE AMERICAN

MINERALOGIST,

VOL. 56, SEPTEMBER_OCTOBER, 1971

POIKILITIC ALBITE IN THE MICROCLINE OF GRANITIC PEGMATITES1 D. Jenolm Frsunn, Departmentof Geology, Arizona State University, Tempe,Arizona 85281
ABSTRACT -small anhedrai crystals of albite occur as unoriented and apparently corroded units enclosed in the large perthitic microcline crystals of some granite pegmatites. Such albite is called loikiJitic albite (PA) because the resulting texture is clearly poikilitic; in similar fashion one may speak ol the perth,i.ticalbite that occurs in microcline perthite. Textural evidence demonstrates that the PA crystals formed more or less contemporaneously with the host microcline crystals, and in some cases definitely earlier. In several deposits PA is cut by later perthitic albite (vein type). Thus not all of the latter can have been formed b_va simole, relatively static unmixing process; significant transportation of perthitic albite material has occurred. These relationships deny the statement by Laves and Soldatos (1963) that the PA is developed by the action of later Na-bearing solutions.

IrvrnooucrroN From an extensivestudy of pegmatite perthites (27 out of 28 samples) made by single-crystal (precession) X-ray techniques,Laves and Soldatos (1963) interpret the relation of the perthitic albite to the host K-feldspar. This involves whether the latter was monoclinic or triclinic at the time of exsolution or recrystallization. They arrive at the following sequenceof five stages: f. Formation of large monoclinic crystals containing Na-feldspar dissolved in K-feldspar. II. With sufficiently slow cooling, unmixing yields chiefly macroperthite (vein or braid) from the monoclinic sanidine host. The growth surfacesof the albite are roughly parallel to [c]. It is possible that some microperthite (string) forms at this stage, and in some casesas this continues to {orm it may develop into vein perthite. III. With more cooling, at about 500oC,the K-feldspar is converted to "orthoclase," a submicroscopically-twinned material containing triclinic domains. At this stage microperthite (film) is produced, as well as Type I cryptoperthite (perthite from the monoclinic state).
I This paper was submitted to another journal in August 1968 while the writer was still at the University of Chicago. It lay buried in Europe 22 months without efiective action being taken on it. ft was finally returned to the writer only after the threat that extteme measures would be taken to recover it. I am grateful to Philip II. Abelson, Bditor of Scimce, for helping me.in this matter, and to Richa.rd H. Jahns, Stanford University, for suggestions , : Ieading to improvemenl" in this presentation.

t769

1770

D. JEROME FISHER

IV. With continued cooling the K-feldspar becomes"more triclinic," and cryptoperthite Type II (perthite from the triclinic state) may develop. V. In a final stage plate perthite (Laves and Soldatos, 1962) nay form by recrystallization. No attempt will be made here to repeat the arguments which Laves and Soldatosuse to support their theory, but I admit that many of them are cogent. They regard their five stagesas the "normal case" of perthite development. But they recognizethat replacement phenomena in many casesmay modify the results of the unmixing described by them. They list two such possibilities regarded as limits or extremes: (a) Albitization by Na-ions, chiefly from outside sources,replacing the K-ions with substantially no change in orientation of the AlSfuO8structure formed originally as described in Stage I. This process leads to patch perthite and so-calledchessboard albite. Many intermediate states and degreesof replacement may occur. (b) Formation of "cleavelandite." While the albitization described in a) can proceed to varying degreeswithout the help of "solutions," the development of "cleavelandite" appearsto depend on the action of Na-bearing solutions. Laves and Soldatos state that this process is impressively verified by the observations of Fisher (1942, 1945), Adamson (1942), and Andersen (1928). Potrrr,rrrc Arsrrn The remarks in this paper are primarily concernedwith the statement made in (b) above. First let me make clear that the use of the term "cleavelandite" as employed there is not desirable.Typical cleavelandite, which is so commonly present in any but the simplest granitic pegmatite, consistsof albite growing in near-parallel platy massesof definite megascopicsize.It commonly showscurved lamellae that are generally twinned according to the albite law (Fisher, 1968). Distinctly different are the small, isolated, unoriented crystals called "cleavelandite" by Laves and Soldatos; I prefer to designate these poikiliti,c al,bitet(PA), because the resulting texture is clearly poikilitic (Fisher, 1945). Admittedly this is a minor point, since these PA crystals are commonly platy or almost acicular and generally shoy albite twinning; some occur in bunched-up masses,but because they are unoriented rather than in near-parallel growths even these are really not the same as typical cleavelandite. r Lavesand Soldatos (1963, 93) describe kindsof poikiliticalbite.I havenot p. two poikilite." What they designate "cleavelandite," observed tleir "isometric as showing albite twinningasin their Fig. 8, in thepresent paperis translated PA (poikilitic albite) as

POIKII.ITIC

ALBITE

r771

The statement in (b) above cannot be properly attributed to the three cited authors.2 None of these investigators considers the PA to have formed by the action of Na-bearing solutions, but instead they regard it as a direct precipitate from the pegmatite "magma" in its early stages, with many or all of the crystals antedating the formation of the vein perthitic albite. Crystals of PA up to about 5 mm in length are known in the K-feldspar (microcline) crystals of granitic pegmatites in three continents. Excellent examples have been described by Laves and Soldatos (1963) as well as the three authors mentioned in paragraph (b) above. Photomicrographs are given by Rogers and Kerr (1942, Fig. 209) and by Drescher-Kaden (1948, Fig. 59; from Southbury, notLouthboury, Conn.; see Cameron el al.,1954, p. 341). Drescher-Kadenstates (basep. 75): Plagioklasesvom Ebenso zweifelsfrei sind Teile des eingeschlossenen (oder von den ihm voraufgehendenLiisungen) angefressen Bandperthit und abgelaugt. inclusion And in the caption of his figure he refers to the "older plagioclase in the microcline." CnBssnolno Ar.nrru What I regard as the clinching argument in this matter is outlined in an earlier report (Fischer, 1945). This publication is relatively inaccessible; hence the gist of the pertinent statements is ofiered here. High up on the north wall of the High Climb mine in the Black Hills (Fisher, 1942; Sheridan, 1955) cleavage blocks nearly 30 cm long have been broken ofi from much larger single crystals of perthitic microcline. These crystals are cut by coarse,veinlike massesof quartz-albite3 rock. Along contacts with such veins the microcline is completely albitized for a thickness of two to four cm. The albite and microcline are in parallel position as shown by the continuity of the basal cleavageacrossthe faces of specimens.A thin section 7|X9 cm was prepared parallel to the basal cleavageof such a specimen,and from it the photomacrograph shown as Figure 1A was made. Four photomicrographs of selected areas in this section appear as Figures 1B-E. It should be realized that a vein-like mass of quartz and albite formerly extended roughly parallel to (010) just below the base (long side) of Figure 1A. It is abundantly clear from the field and thin-section evidence that a 2In givingad abstract the Lavesand Soldatos erroneous (1963) paper,this same of (1967). preserved Tr<iger in has interpretation been 3 Earliertermed this However albiteprobquartz-cleavelandite by Fisher(1942). rock to Smsustrittu (Fisher,1968). ably shouldnot bereferred ascleavelandite,

1772

D. JEROME FISHER

& s
ati

s
fi qf.

"-i*'l -rl*

PO]KILITIC

ALBITE

r773

lEl

1774

D. JEROME FISEER

"Na-front" invaded the microcline, replaced its K, and yielded this beautiful example of chessboardalbite This was simply a type of ionexchangereaction involving these two alkali metals, with at most only very slight changesin the AlSiaO3-framework.Some muscovite is present (Fig. 1E), but there is no evidencethat it was produced with the aid of the K released in this reaction; certainly this specimen does not show nearly enough muscovite to account for all this K. The albitization took place starting from the base of Figure 1A and extending upward. rt proceededalong surfaces or incipient fissureswhose traceswere approximately the short side of the figure, producing material like vein perthite. As these albite laminae were extended they broadened somewhat, and where closely-enoughspaced they gradually coalescedto form patch perthite. Along their surfaces of juncture the twinning lamellae are offset, giving the typical chessboardtexture. The extremely sinuous character of the boundary between the chessboard albite and the perthitic microcline is made clear in Figures 1A, B, and C. These views also show how individual units of the chessboardalbite tend to become coarserwith increasingdistance from the microcline. But the outstanding feature relative to the present discussionis the presence of PA in all four of the photomicrographs. It is most pronounced in Figure 1D, less so in Figures 18, E and only two small crystals are visible along the left edge of Figure 1C. The PA has two outstanding characteristics: 1) the crystals are diversely oriented; i.e., poikilitic within the host, with no tendency to line up with any other feldspar, and 2) they are anhedral, and generally appear to be corroded or partially resorbed. The perthitic albite, the chessboard albite, and the PA all appear to be of essentially the same composition, An 0-5, which is as closeas one can determine the composition of albite by optical (U-stage) work, according to the data in Burri et al. (1967). f consider it highly unlikely that the occurrenceof albite in these three different manifestations can be related to any significant compositional differences.Albite which replacesmicrocline tends to exist in parallel orientation to it; in fact this is the main criterion for assuming that replacement has occurred. Such albite should not be confused with that PA, or with polycrystalline albite of a later generation that may be fairly common as veins cutting acrossthe perthitic microcline. The relationships shown in Figure 1D indicate that the PA did not replace the chessboard albite. Such PA is common in the perthitic microcline of the pegmatities of the Black Hills and other regions, as is documented herein. A dozen or more PA crystals are present in the upper part of Figure 1A completely surrounded by the perthitic microcline, hence they must have been present in the perthitic microcline

POIKILIT]C ALBITE

1775

which was replaced by the chessboardalbite. It must be concluded that, contrary to the rather involved explanationof Laves and Soldatos(1963), such crystals represent one of the earliest of pegmatite minerals in the Kfeldspar unit. Sheridan (1955,p. 87) notes that while "the albite of the matrix appears to replace microcline-perthite in some specimens, in others the microcline replaces and veins the albite of the matrix and poikilitically includes crystals of 'cleavelandite'," but he says nothing regarding the time of formation of these small poikilitic "cleavelandite" crystals. Exaupr-Bs Becausethe large thin-section from which Figure 1A was made is necessarily of greater than standard thickness, a normal thin section was ground from the same sample and from it the photomicrographs of Figures 2-4 were prepared. The PA crystals shown in Figures 28 and 38 were studied on the U-stage and found to be of composition An 0-5. The crystal shown in Figure 28 is a dual twin after the Carlsbad A law, and that in Figure 38 is polysynthetic according to the albite law. These photomicrographs reveal the anhedral, even corroded, borders of the PA crystals. The discontinuity of individual units of the chessboardalbite twins at their contacts with the PA or the perthitic microcline is clearly shown in Figures 3B and 4, fortifying the thesisthat the chessboard albite is younger than either of the other two feldspars. The relationships shown in Figure 28 are highly incompatible with the notion that the PA was formed by replacement of the perthitic microcline: why should the introduction of Na (or indeed of all the components of albite) cause the replacement of the microcline Iattice in every possible orientation except the one that would involve a minimum rearrangement of the AISLO8groups? Figure 3 shows the only observedcasewhere a PA crystal is cut in two by the chessboardalbite. Note that the Iarger remnant (on the right) is entirely surrounded by chessboardalbite, with the traces of the respective (010) planes nearly perpendicular to each other, whereas the smaller remnant projects into the perthitic microcline. It is clear that the individual units of the chessboardalbite are interrupted at the PA except along the tiny tip at the left end of the larger PA remnant. This kind of discontinuity, indicating that the chessboard albite on opposite sidesof a remnant of PA or of perthitic microcline grew from different directions and so from different seeds, also is very clear in Figure 4. The lower side of the PA crystal in Figure 48 is spangledwith shreds of muscovite, and to the right of this crystal are two linear areas of muscovite shreds that may mark the locations of former PA individuals. However it should be em-

t776

D, JEROME FISHER

Frc. 2. Photomicrographs at two scales (A and B) from a standard-size thin-section ground from the same sample as used for Fig. 1. Shows a poikilitic albite (PA) crystal Iargely in a microcline remnant, but with perthitic albite cutting across its left end. The black band along the lower rim of the PA crystal in Fig. 28 seems to be apatite.

POIKILITIC

ALBITE

I'rc. 3. Photomicrographs at two scales (A and B) from a standard-size thin-section ground from the same sample as used for Fig. 1. Shows a PA crystal cut in two by chessboard aibite; see text for detailed information.

1778

D. JEROME F]SHER

(b) Frc. 4. Two photomicrographs from a stanrlard-size thin section ground from the same sample as used for l-ig. 1. Fig. 4a shows a remnant of perthitic microcline in chessboard albite. Along the middle of the bottom edge appears the right end of the PA crystal rvhich is shown in the left-center of Fig. 4b. In this view note how the individuals of the chessboard albite are interrupted bv the PA. See further discussion in the text.

POIKILITIC ALBITE

1779

phasizedthat the associationof shreddy muscovite with the borders of the PA crystals (note for example the two crystals crossing the upper right and lower left corners of Fig. 28; see Fig. 2A) is the exceptional, not the ordinary, occurrence. Figure 5A shows the only observed casewhere a PA individual (nearly horizontal acrossthe middle of the view) is divided into two units through partial replacement by microcline. Here one of the PA remnants is very slightly turned from parallel orientation with the other, but the twinning matches up so nicely that there is no doubt that they once constituted a single (twinned) crystal. When looking at this picture one is strongly reminded of Adamson (1942, Fig. 2 on PI. IV). The Iower of the two nearly parallel crystals of PA that extend horizontally acrossthe middle of Figure 58 contains four vein-like stringers of perthitic albite projecting halfway acrossit, Figure 6 shows corrosion and veining of PA by perthitic albite. This PA is unusual in that it is twinned according to three laws: albite (the polysynthetic twinning in the lower half of the crystal shown on the right side of Fig. 68), Roc Tourn6 (the two units shown on the left side of Fig. 6B), and Carlsbad (ttre ligtrt unit in the lower part of the right end of the crystal of Fig. 68 along with its upper part). Both Figure 6 and Figure 7 representa sample of perthitic feldspar from the BeecherNo. 2 pegmatite (Longview mine) in the Black Hills (Fisher, 1942; Page et ol. 1953). The PA crystal in Figure 7, twinned on the albite law, is not only cut substantially in two by perthitic albite, but its right-hand part is slightly rotated from parallelism with the main left-hand portion. The right edge of the remnant on the right is corroded by another perthitic albite vein. I have six photomicrographs of microclines from the Custer Mt. Lode (Fisher, 1945; Page et aI., 1953) all showing excellent examples of anhedral PA cut by perthitic albite, but here only one is reproducedas Figure 8. This PA is a polysynthetic twin (albite law) that has been corroded and completely severed by the perthitic albite. Clearly the crystal was not broken in two and the two parts shoved away from each other, because the adjacent ends of the two parts do not match. Further, there is no indication of bending or other deformation in the PA. And it seems hardly reasonable to assume that microcline first cut through the PA unit, and that it was later completely replaced by perthitic albite. The perthitic albite cutting the upper part of the PA crystal comprises two units, and the two sets of albite twins are offset from each other as is typical of chessboardalbite. Andersen(1928,p. 153) notes that very often the PA is penetratedby albite of the vein perthite; thus the vein perthite cannot have been formed by exsolution in the solid state, but must have grown from circu-

1780

D. JEROME FISEER

Frc. 5. Photomicrographs from a perthitic microciine crystal from the lIigh Climb Lode. Fig. 5a shows a PA crystal split into two corroded units by the microcline. The PA crystal of Fig. 5b is cut by four veinlike masses of perthitic albite extending from the perthitic microcline.

POIKILITIC ALBITE

(b) Frc. 6a. Microcline (light gray) with its perthitic albite (white) cutting across an anhedral crystal of PA twinned according to albite, Carlsbad, and Roc Tourn6 laws. Fig. 6b shows the enlarged right center part of Fig. 6A Beecher no. 2 pegmatite, Black Hills.

1782

D. TEROME FISHER

(b) I-rc. 7. Photomicrographs at fir'o scales (a and b) showing relationships microcline and two generations of albite. Beecher no. 2 pegmatite, Black Hills.

POIKILTTIC ALBITE

r783

Frc. 8. Perthitic microcline enclosing a corroded crystal of PA that has been cut in two b-v perthitic albite. Custer Mt. Lode, Black Hills.

Iating solutions. Adamson (1942, p. 31) agrees.However it is clear that while this perthitic albite as a result of exsolution may develop initially as tiny specks evenly distrjbuted throughout the microcline, such a condition is not observed during thin-section study. fnstead the perthitic albite has aggregated into sheets or vein-like massesthrough diffusion in the solid state. Thus presumably some transportation of the materlal of the perthitic albite must have taken place. When perthites which have been homogenizedby heating as dry melts are allowed to cool, no exsolution is observed.But according to Richard H. Jahns (personal communication) subsolidus difiusion can occur rapidly over large domains under Iaboratory conditions if water under pressure is present. Relationships very similar to those here describedhave been achieved experimentally in only five days. In the description of the Hugo pegmatite (Norton et al., 1962), the presenceof PA in perthite is noted for intermediate zones 3 and 5 (see their fig. 30), and possibly in 6. There also appears to be a crystal of PA in the center of the photomicrograph from zone 5 perthite shown by Orville (1967,Fig. 1). Through the kindness of ProfessorT. F. W. Barth, I was able to study one of Andersen's thin sections (A 23-714a) from which he prepared the photomicrograph shown as his Plate VI, Figure 1 (Andersen, 1928E. An

1784

D. TEROME FIS]IER

Frc. 9 Perthitic albite invading a corroded crystal of PA in microcline. Tangen, Norway.

Frc 10. Corroded poikiiitic albite crystal surrounded by microcline. Tangen, Nonvay.

POIKILITIC ALBITE

1785

enlargedview of the PA crystal that appearsabove the center of his figure is shown here as Figure 9. This corroded crystal (a dual albite twin) lies within microcline and has been invaded by a mass of perthitic albite. Similar relationships appear elsewherein this same thin-section, which also yielded the photomicrograph reproduced as Figure 10. Here a corroded PA crystal of unusual appearance is completely surrounded by microcline which Iacks pericliae twinning in the lower part of the view. The lower two-thirds of the PA crystal is a polysynthetic albite twin, whereas the two major parts in its upper half constitute a Carlsbad twin. The dark portion at the top of the crystal and the nearly parallel dark portions in its lower half form a Roc Tourn6 twin. CoNcr,usroxs In summary, the explanation by Laves and Soldatos (1963) that the PA in granitic pegmatites was introduced by late-stage Na-bearing solutions is here regarded as untenable. They state: "Na-bearing solutions may enter into the perthitic microcline leading to the solution of K-feldspar and the deposition of Na-feldspar. Since the-.K-feldspar is especially unstable in the presence of Na-bearing solutions, it is replaced, whereas there is no particular driving force present to cause the replacement of the Na-feldspar of the vein perthite by the Na-feldspar of the PA. Therefore the PA will advance in the microcline, in placesgrowing around the vein perthite present. and the latter will only locally recrystallize in the new PA orientation." Each of these three sentencescontains statements open to question. Why must "solution of K-feldspar" be inferred? Why not postulate simply an ion-exchangetype of reaction with Na replacing K? Regarding the last part of the secondsentence,Na-feldspar apparently doesreplace Na-feldspar, or at least causes to assumea new orientation, as shown by it veins of perthitic albite transecting PA, and by chessboardalbite cutting acrossPA (Fig. 3). But I take particular exception to the final sentence. What actually happens when Na-bearing solutions enter microcline is shown in Figure 1. Not only does the K-feldspar change to Na-feldspar, but the result is non-poikilitic albite-in fact albite in parallel growth, in this casechessboard albite. Emmons and Mann (1953,p. 47) note that in Wisconsin granites where microcline is replaced by plagioclase"it is normal for the ensuing plagioclaseto inherit the albite twin lamellae of the microcline almost lamella for lamella." Laves and Soldatos(1963) go on to say: "Andersen noted that the PA crystals are often not bounded by sharp crystal faces, but rather appear to be eaten into or corroded

1786

D. TEROMEFISHER

by the microcline.To this it should be stated that one cannot understand how, during the albitization by vein perthite postulated by Andersen, a development of microcline at the expense of the PA could have taken place." I would add here that I have never seena crystal of true PA that was euhedral or even definitely subhedral. If, during the consolidation of a "pegmatite magma", one of the first things to happen during the major microcline stage was the formation of a "shower" of tiny albites (an entirely reasonablehypothesis in terms of what is now known of the physical chemistry of such a system), then these crystals would become engulfed during the development of the much larger microcline crystals. Under these conditions these tiny albites could be expected to 1) suffer some corrosion,and 2) be swallowedpoikilitically by the growing microcline. AII of this must have occurred while the temperature was such that little or none of the perthitic albite had as yet becomeunmixed. If Nasolutionswere replacingmicroclinewith albite, why should they deposit a little Na-feldspar here and there at widely-spaced distances (considering the tiny sizesof the PA crystals), and with new AlSLOe-structure orientations, rather than building up a non-poikilitic mass near the place Andersen's views are not in agreementwith the statement in of entrance? the last sentenceof the foregoing quotation. Laves and Soldatos(1963) continue: "W'e are rather inclined to consider the'corroded'edges of the PA crystals to be merely apparent, and we think they can be better explained as follows: the growth of the PA proceededmost easily where the microcline was most disturbed. The latter was especiallythe case where the microclineshows 'diffuse' extinction; that is, where there are diffuse twin boundaries; i.e., regions where a disordered and therefore unstable condition existed with respect to the AI/Si distribution." The corroded nature of the PA crystals seemsobvious to me, and this characteristic is well shown by the photomicrographs. It also should be noted that if the PA crystals were engulfed by growing giant microcline crystals, thesehosts may have sufferedminor and local attacks of indigestion resulting in disturbed border areas. While they may have resorbed some of the PA crystals, clearly there were many that they couldn't finish offl
RrrnntNcns Aortusox, O. J. Q942) Minerals of the Varutriisk pegmatite XXXI. The feldspar group. tr'iir. Fiirh.64, 19-54. GeoI.

POIKILITIC

ALBITI'

1781

'fhe AunnnseN, O. (1928) genesis of some types of feldspar from granite pegmatites. /forsft Ceol,.Tid.s.l0, 113-206. Bunu, C., R. L. Panxrn, E WnNr (1967) Die opti.scheOrientierung der P!'agiokl'ase.Birlr.hziuser Verlag, Basel und Stuttgart, 334 pp. CeuenoN, E. N., el aI. (1954) Pegmatite Investigations 1942-45, Ne'w England t/. S. G e o l .S u r t t .P r o J . P a p . 2 5 5 , 3 5 2 p p . Dnnscnop-K.ronN, F. K. (.1948) Die Feld.spat-quarz-reaktionsgeJiige Gronite uwJ Gneise. der Springer Verlag, Berlin, 259 pp. Erruous, R. C., el al,. (1953) Selected petrogenic relationships of plagioclase. Geol,.Soc. A rner. M em. 52, 142 pp. Frsnnn, D. I . 0942) Pegmatites of the Custer district. S. Dak. Geol,. Sum. Rept. Intsest. 44, 35 pp. - -- (1945) Mineralogy of some pegmatites near Custer .S. Dala. Geol. Sutn. Rept. Inttest. 50, 92 pp. -(1968) Albite var. cleavelandite, and the signs of its optic directions. Amer. Mineratr.53,1568-1578. Lavns, l'., exo K. Soloeros (.1962) Plate perthite, a new perthitic intergrowth in microclire single crystais, a recr)'stallization product. Z Krisl,al,logr. ll7r218-226. -- -(1963) Die Albit,/mikroklin-orientierungs-beziehungen in Mikroklin-perthiten und deren genetischeDeutung. Z. Kr,istdlogr.1l8, 69-102. N o R r o N , I . J . , L . R . P l c r , A N D D . A . B n o s r ( 1 9 6 2 )G e o l o g y o f t h e H u g o p e g m a t i t e , K e y stone, South Dakota. U. S . Geol. S wa. P r ol. P ap. 297 -8, 49-727 . Onvrr-r,r, Prlr.r.rp M. (1967) Unit-cell parameters of the microcline-low albite and the sanidine-high albite solid solution seties. Amer. Mi.neral.52, 55-86. Pecn, L. R., et aL (1953) Pegmatite investigations 1942-45, Black Hills, U.S. Gnl'. Sun. ProJ.Pap.247,228 pp. RocERS, A. F., .twn P. F. Krnn (1942) Optical Mineral,ogy. McGraw-Hill Book Co., New York, 390 pp. SunntnAN, D. M. (195.5) Geology of the High Clinrb pegmatite, Custer County, South Dakota. Lt.S. Geol.Sru.u.Bull.l015-C, 59-98. Ttcicen, W. E. (1967) Optische Butirnmung d,er gesteins-bi,ldend.en Minerale, Teil, 2 Terl,band (p.708). Schweizerbart Verlag, Stuttgart, 822 pp. M anwscript recei:t,ed. August 11, 1970; accepted. publi,coti.on,ilIareh 31 , 1971 . lor